Effect of Shielding Gas Composition on Phase ... · PDF fileIn this work, to elucidate the eﬀect of shielding gas composition on the phase transformation and the mechanism of pitting

Effect of Shielding Gas Composition on Phase Transformation and Mechanism

of Pitting Corrosion of Hyper Duplex Stainless Steel Welds

Seok-Hwan Jang, Soon-Tae Kim*, In-Sung Lee and Yong-Soo Park

Department of Materials Science and Engineering, Yonsei University, Seoul 120-749, Korea

The effect of shielding gas composition on the phase transformation and the mechanism of pitting corrosion of hyper duplex stainless steel(HDSS) welds was investigated in highly concentrated chloride environments. The resistance to pitting corrosion of a HDSS tube after weldingwith Ar shielding gas supplemented with N2 was increased due to a decrease of the PREN (Pitting Resistance Equivalent Number) differencebetween the �-phase and the �-phase in the weld metal and the heat affected zone. Cr nitrides (Cr2N) were precipitated in the weld metal and theheat affected zone due to a high �-phase content. Cr-depleted zone adjacent to Cr2N decreased the resistance to pitting corrosion.[doi:10.2320/matertrans.M2010414]

(Received December 6, 2010; Accepted March 11, 2011; Published May 18, 2011)

Keywords: shielding gas, polarization, pitting corrosion, transmission electron microscope, welds

1. Introduction

Duplex stainless steels with nearly equal fraction of ferrite(�) phase and austenite (�) phase are increasingly being usedfor various applications such as power plants, desalinationfacilities, the off-shore petroleum industry, and chemicalplants due to their high resistance to stress corrosion crackingand pitting corrosion, good weldability, excellent mechanicalproperties and relatively low cost owing to the addition oflow Ni, as compared with austenite stainless steels.1–3)

In general, it is well known that super duplex stainlesssteels such as UNS S32750, UNS S32760 and UNS S32550are defined as duplex stainless steels with a PREN (PittingResistance Equivalent Number (PREN) = [mass% Cr] +3.3([mass% Mo] + 0.5 [mass% W]) + 16 mass% N4,5)) of40�45. Hyper duplex stainless steel (HDSS) such as UNSS32707 is defined as a highly alloyed duplex stainless steelwith a PREN in excess of 45.4) The corrosion resistances ofduplex stainless steels are determined by the fraction of the�-phase and �-phase and by the Cr and Mo depleted zoneadjacent to the secondary phases such as detrimentalintermetallic phases (�, �), carbides, and nitrides.

Hence, when duplex stainless steels are welded, the heatinput6–11) and incorporation of nitrogen in the shieldinggas12–14) can significantly affect the resistance to pittingcorrosion of duplex stainless steels. Control over the balanceof the �-phase and �-phase in the weld metal (WM) and theheat affected zone (HAZ) is important from a corrosionviewpoint, because the corrosion resistance deteriorates witha high content of the �-phase. Also, the content of the �-phase and �-phase is important for fracture toughness. As thecontent of the �-phase in duplex stainless steel increases, theimpact toughness decreases. Therefore, a proper balance ofthe �-phase and �-phase must be maintained. For a givenplate thickness, the cooling rate decreases with an increase inthe heat input. Also, for a given heat input, the cooling ratedecreases as the plate thickness decreases. For these reasons,the welding heat input cannot be considered in isolation.

For the following discussion, the thickness of the rolledplate and joint configuration is assumed to be the same. Thecontent of the �-phase in the duplex stainless steel is afunction of the heat input and cooling rate: the lower the heatinput, the higher the ferrite content and the lower the impacttoughness.15–22) Draugelates et al.17) reported that highercooling rates suppress the diffusion-controlled processes inaustenite reformation. Hence, the original phase ratio (1 : 1 invol%) of the �-phase to the �-phase is shifted towards ahigher content of the �-phase.

The secondary phase precipitation is also significantlyaffected by high cooling rates. Lippold et al.20) and Kirievaet al.21) reported that the presence of chromium-rich nitrides(Cr2N) is observed over a wide range of cooling rates and theeffect is particularly evident for microstructures with a highferrite content, which are usually the result of fast coolingrates. These chromium rich nitrides also significantly de-crease the impact toughness and pitting corrosion resistance.A risk of Cr2N formation in the �-phase is also noted with anincrease in the �-phase and increased nitrogen levels due tothe lower solubility of nitrogen in the �-phase. However, highcooling rates do reduce �-phase or �-phase precipitation.

Nitrogen is a strong �-stabilizer, and it increases thetemperature at which the transformation of the �-phase to the�-phase occurs.6,23,24) N has also been found to accelerate thepartial transformation from the �-phase to the �-phase duringcooling after welding.6) In addition, it may help in bringingabout homogenization of the Cr distribution in these twophases.25–27) Appropriate control over these factors can resultin a significant improvement in the integrity of the weldedassembly and, in particular, its corrosion resistance. Giventhe above considerations, a mixed gas of Ar and N2 as ashielding gas has recently been used to absorb the nitrogeninto the WM of the duplex stainless steels.28–31)

Accordingly, it is important to quantitatively verify themechanisms underlying the effect of nitrogen in the shieldinggas on the corrosion behavior of the Cr-depleted zoneadjacent to Cr2N and the difference in the corrosionresistance between the �-phase and �-phase in the weldmetal (WM) and heat affected zone (HAZ) of the HDSS.*Corresponding author, E-mail: [email protected]

Materials Transactions, Vol. 52, No. 6 (2011) pp. 1228 to 1236#2011 The Japan Institute of Metals

http://dx.doi.org/10.2320/matertrans.M2010414

In this work, to elucidate the effect of shielding gascomposition on the phase transformation and the mechanismof pitting corrosion of hyper duplex stainless steel (HDSS)welds, a metallographic examination, a potentiodynamicpolarization test, a critical pitting temperature test, scanningelectron microscope and energy dispersive spectroscope(SEM–EDS) analyses, a scanning Auger multi-probes (SAM)analysis of �-phase and �-phase, and a transmission electronmicroscope (TEM) analysis of Cr2N precipitates were carriedout.

2. Experimental Procedures

2.1 Calculation of phase diagram and the equilibriumfractions of each phase

The phase diagram and the equilibrium fractions of eachphase were calculated against the temperature for the HDSSalloy using a commercial Thermo-Calc software package.

2.2 Material and heat treatmentHDSS tubes with an outer diameter of 19.05 mm and

thickness of 1.5 mm were manufactured using the gastungsten arc welding (GTAW) method with a heat input of80 KJ m�1 and a shielding gas composed of Ar-5 vol% N2 andAr. The chemical compositions of the HDSS tube with PREN51 and a commercial super duplex stainless steel tube-SAF2507 (UNS S32750) with PREN 46 are presented in Table 1.

2.3 Microstructural characterizationTo observe the optical microstructures of the weld metal

(WM), heat affected zone (HAZ), and base metal (BM) in theHDSS welds, the tubes were ground to 2000 grit using SiCabrasive papers, polished with diamond paste, and thenelectrolytically etched using 10 mass% KOH. The chemicalcompositions of the �-phase and �-phase were analyzedusing a SEM and an EDS attached to a SEM. The nitrogencontent was analyzed using a SAM.

The Cr2N precipitates formed in the HDSS tube during theGTAW were analyzed by both a carbon replica technique anda thin film technique with a transmission electron microscopy(TEM). The carbon replicas were prepared by electrolyticallyetching the polished samples with an etchant of 10 mass%KOH. The thin foil specimens were prepared electrolyticallyat the inter-electrode voltage of 25 V in 10 mass% perchloricacid plus 90 mass% methanol.

2.4 Corrosion testsTo analyze the effect of the incorporation of nitrogen in the

shielding gas on the resistance to pitting corrosion of theHDSS welds, the potentiodynamic anodic polarization testswere conducted. To measure the pitting potential (Ep) and thepassive current density (Ip) of the alloys, the test was

conducted in the same volume mixture of deaerated 0.5 NHCl aqueous solution and deaerated 1 N NaCl aqueoussolution at 328 K and a deaerated 25 mass% NaCl at 333 Kaccording to ASTM G 5.32) The test specimens were joinedvia soldering with copper wire (95 mass% Sn-5 mass% Sb),and then mounted with an epoxy resin. One side of thesample was ground to 600 grit using SiC abrasion paper.After defining the exposed area of the test specimen as0:5� 10�4 m2, the remainder was painted with a transparentlacquer. The potentiodynamic test was carried out using thewelded zone (WM+HAZ+BM) with an exposed area of0:5� 10�4 m2. The test was conducted at a potential rangeof from �0:65 V to +1.1 V vs. SCE (saturated calomelelectrode) and at a scanning rate of 1� 10�3 V s�1, using aSCE.

A critical pitting temperature (CPT) test was conducted in10 mass% aqueous solution of FeCl3�6H2O with a pH level of0 according to ASTM G48 method A.33) The specimens wereground to 100 grit using SiC abrasion paper. The solutiontemperature was increased by 278 K every 259.2 ks. After thetest was completed, the corrosion products were removedusing ultra-sonic equipment in acetone. Pitting corrosion isconsidered to be present if the local attack is 2:5� 10�5 m ormore in depth. An optical microscope was used to observethe initiation and propagation of the pitting corrosion aftermeasuring the CPT.

3. Results and Discussion

3.1 Calculation of the phase diagram and equilibriumfractions of each phase

The phase diagram of the HDSS alloy was calculated usinga commercial Thermo-Calc software package, and provides aroadmap of the metallurgical behavior (Fig. 1(a)). A sec-tional view of 27 mass% Cr illustrates that the alloy solidifiesprimarily as an �-phase, and some of the �-phase transformsto �-phase with a decrease in the temperature. As thetemperature decreases further, the �-phase decomposes into�-phase (�) and secondary austenite (�2) according to theeutectoid reaction.

L! Lþ �! Lþ �þ � ! �þ �! �þ �2 þ �! �2 þ �

ð1Þ

The equilibrium fractions of each phase against thetemperature for the HDSS alloy were calculated using acommercial Thermo-Calc software (Fig. 1(b)) package.It was predicted that the content of 50 vol% �-phase and50 vol% �-phase is obtained about 1363 K.

3.2 Effect of shielding gas composition on the micro-structures of the HDSS welds

Figure 2 shows the effect of the shielding gas composition

Table 1 Chemical composition of hyper duplex stainless steel and a commercial super duplex stainless steel-SAF 2507 (mass%).

Alloys C Cr Ni Mo W N Mn S Ce La Ba Si PREN�3

HDSS�1 0.02 27.0 7.23 2.57 3.23 0.34 1.96 0.001 0.0137 0.0045 0.0006 0.20 51.0

SAF 2507�2 0.013 24.8 6.84 3.90 — 0.26 0.55 0.002 — — — 0.32 45.5

�1The developed alloy, �2the commercial alloy: SAF 2507 (UNS S32750), �3PREN ¼ [mass% Cr]þ 3:3ð½mass% Mo� þ 0:5½mass% W�Þ þ 30½mass% N�

Effect of Shielding Gas Composition on Phase Transformation and Mechanism of Pitting Corrosion of Hyper Duplex Stainless Steel Welds 1229

on the microstructures of the HDSS welds. The �-phasesformed in the BM were elongated in the direction of the hotrolling, irrespective of the chemical compositions of the

shielding gas. The morphology of the �-phases formed in thegrain boundaries of the �-phases in the WM and the HAZ wasirregular. The �-phases in the �-grains of the HAZ wereformed fewer than those of the WM.

Figure 3 presents the effect of the shielding gas compo-sition on the ferrite content of the HDSS welds. The contentof the �-phase in the WM of the HDSS tube being weldedwith Ar-5 vol% N2 was higher than that of the HDSS tubebeing welded with Ar. The addition of N2 as a strong �-stabilizer to the Ar shielding gas in the GTAW process hadpositive influences on the prevention of nitrogen loss and alsoincreased the temperature required for the transformationfrom the �-phase to the �-phase, thereby promoting thetransformation from the �-phase to the �-phase during thecooling period after welding.

Figure 4 shows the results of the TEM analysis of the Cr2Nformed in the WM of the HDSS welds after welding with Ar-5 vol% N2. All precipitates observed in the WM were rod-like Cr nitrides-Cr2N (Figs. 4(a) and (b)). Neither carbidesnor intermetallic compounds such as the �-phase and �-phasewere detected in the WM. Based on the analysis of thediffraction pattern in Fig. 4(c), Cr2N has a hexagonal closepacked (HCP) structure. A Cr-depleted zone, which decreas-es the corrosion resistance adjacent to the Cr2N, was

(a)

(b)

Fig. 1 (a) the phase diagram and (b) the equilibrium fractions of each

phase for the HDSS alloy using a Thermo-Calc software package.

* Base Metal: BM, Weld Metal: WM, Heat Affected Zone: HAZ

Fig. 2 Effect of shielding gas composition on the optical microstructures of the HDSS welds: (a) welded with Ar and (b) welded with Ar-

5 vol% N2.

0

5

10

15

20

25

30

35

40

45

50

55

60

65

Fer

rite

vo

lum

e p

erce

nta

ge

(vo

l.%)

WMHAZBM

HDSS : ArHDSS : Ar-5 vol. % N2

Fig. 3 Effect of shielding gas composition on the ferrite contents of the

HDSS welds.

1230 S.-H. Jang, S.-T. Kim, I.-S. Lee and Y.-S. Park

observed (Fig. 4(d)). As the content of the �-phase in theHAZ is much larger than that in the WM (Fig. 3), Cr2N canprecipitate in the HAZ. Meanwhile, as the content of the �-phase in the WM and HAZ of the tube being welded with Aris much larger than that of the tube welded with Ar-5 vol%N2 (Fig. 3), Cr2N also can be precipitated in the WM andHAZ of the tube welded with Ar.

In summary, when a HDSS tube with a small thickness of1:5� 10�3 m is welded with a low heat input of 80 KJ m�1,the cooling rate of the tube is much faster than that of a thicktube welded with a high heat input. The content of the �-phase and Cr2N in the WM and HAZ, which are stable at hightemperature, during cooling after welding using the GTAWmethod increased due to the low heat input and high coolingrate. In contrast, intermetallic compounds such as the �-phaseand �-phase, which deteriorate the corrosion resistance andthe impact toughness, were not precipitated.

3.3 Effect of shielding gas on the resistance to pittingcorrosion

Figure 5(a) shows the effect of the shielding gas compo-sition on the potentiodynamic anodic polarization behaviorof the HDSS welds, HDSS base metal, and a commercialSAF 2507 welds in the same volume mixture of deaerated0.5 N HCl aqueous solution and deaerated 1 N NaCl aqueoussolution at 328 K according to ASTM G5. In general, thepitting potential (Ep) is defined as the breakdown potentialthat destroys a passive film. As the Ep of an alloy increases,the resistance of the alloy to pitting corrosion increases. TheEp of the HDSS welds was approximately equivalent to that

of the HDSS base metal. However, technically, it is not the Ep

because the abrupt increase of anodic current at the pittingpotential range is due to an oxygen evolution. Hence, pittingcorrosion did not occur, irrespective of the chemicalcomposition of shieding gas. However, the passive currentdensity (Ip) of the HDSS welds was decreased from 6:9�10�1 A m�2 to 2:4� 10�1 A m�2 due to the addition of N2 tothe shielding gas. Meanwhile, based upon the measured Ep,the resistance to pitting corrosion of the HDSS welds is foundto be superior to that of a super duplex stainless steels-SAF2507 (Ep, Ar-5 vol% N2

: 0.143 V vs. SCE, Ep, Ar: 0. 048 V vs.SCE), as indicated in Fig. 5(a).

Figure 5(b) shows the effect of the shielding gas compo-sition on the potentiodynamic anodic polarization behaviorof the HDSS welds and base metal in deaerated 25 mass%NaCl aqueous solution at 333 K according to ASTM G5.The abrupt increase of anodic current at the pitting potentialrange of the base metal is due to an oxygen evolution,whereas pitting corrosion occurred in the HDSS welds.However, the Ep of the HDSS tube was increased from0.545 V vs. SCE to 0.670 V vs. SCE due to the addition of N2

to the shielding gas. The Ep (0.670 V vs. SCE) of the HDSStube is much higher than that (0.392 V vs. SCE) of thecommercial SAF 2507 tube under the same shielding gas ofAr-5 vol% N2.

Figure 6 shows the effect of the shielding gas compositionon the critical pitting temperature (CPT) of the HDSS weldsand the commercial tube-SAF 2507 welds. The CPT of theHDSS tube was increased from 338 K to 343 K due to theaddition of N2 to the shielding gas in GTAW process. The

(a)

(c)

-3001920212223242526272829303132333435363738

Con

cent

rati

on o

f C

r (m

ass

%)

Distance, l / nm

(b)

(d)

300200100-100 0-200

Fig. 4 TEM analyses of Cr2N precipitated in the weld metal of the HDSS welds after welding with Ar-5 vol% N2: (a) the morphology of

Cr2N observed by a carbon replica technique, (b) the morphology of Cr2N observed by a thin film technique, (c) the diffraction pattern

and (d) the line analysis of Cr surrounding Cr2N.


CPT (343 K) of the HDSS tube was higher than that (333 K)of a commercial tube-SAF 2507 under the same shielding gasof Ar-5 vol% N2.

Based on the above results of the electrochemical and CPTtests, it was found that the HDSS tube with PREN 51 showeda noticeably better corrosion resistance than a commercialSAF 2507 tube (UNS S32750) with PREN 46 in a highlyconcentrated Cl� environment.

3.4 Mechanism of the effect of the shielding gascomposition on the resistance to pitting corrosionof the HDSS welds

Figure 7 presents the effect of the shielding gas compo-sition on the initiation and propagation of pitting corrosion ofthe HDSS welds after a critical pitting test in 10 mass%aqueous solution of FeCl3�6H2O. Irrespective of the shield-ing gas composition and the zones of the WM and HAZ, the

pitting corrosion was selectively initiated at the �-phase, andthen was propagated from the �-phase to the �-phase.

To clarify the difference in the resistance to pittingcorrosion between the �-phase and the �-phase, the contentof Cr, Mo and W in the �-phase and the �-phase of the HDSSwelds was quantitatively measured using a SEM–EDS andthe nitrogen content was measured using a SAM. Then, thePitting Resistance Equivalent Number (PREN) values of the�-phase and the �-phase were calculated (Fig. 8).

Lorenz and Medawar developed an experimental equationbetween the PREN and the alloy composition as follows.34)

They reported that the PREN of stainless steels, a criterionfor the resistance to pitting corrosion, is defined by thefollowing experimental equation, and is calculated accordingto the chemical compositions (mass%) of major alloyingelements such as Cr, Mo, and N, which greatly affect theresistance to pitting corrosion of stainless steels.

PREN ¼ ½mass% Cr� þ 3:3½mass% Mo� þ 16½mass% N�ð2Þ

Meanwhile, Okamoto observed that W increases theresistance to pitting corrosion of DSS by half that of Mo.5)

PREN ¼ ½mass% Cr� þ 3:3ð½mass% Mo�þ 0:5½mass% W�Þ þ 16½mass% N�

ð3Þ

However, Bernhardsson reported that while the formula isrelevant for austenitic stainless steels and the austenitic phasein duplex stainless steels, it is not valid for duplex stainlesssteels35) for the following reasons. First, nitrogen is nearlycompletely solutionized in the austenite phase in duplexstainless steel whereas it is rarely solutionized in the ferritephase in duplex stainless steel. The solubility of nitrogen inthe ferrite phase in duplex stainless steel has a maximum valueof 0.05 mass%. Second, the addition of nitrogen changes thepartitioning coefficients for chromium and molybdenum.36)

As a result, nitrogen considerably increases the PREN of the�-phase in duplex stainless steels and thus increases theresistance to pitting corrosion of the �-phase, compared withthat of the �-phase. For these reasons, Bernhardsson proposesthat the PREN coefficient for the nitrogen content in duplexstainless steels is roughly doubled, from 16 to 30.

PREN ¼ ½mass% Cr� þ 3:3ð½mass% Mo�þ 0:5½mass% W�Þ þ 30½mass% N�

ð4Þ

(a)

(b)

Fig. 5 Effect of shielding gas composition on the potentiodynamic anodic

polarization behaviors of the HDSS welds, HDSS base metal, and a

commercial SAF 2507 welds: (a) in the same volume mixture of deaerated

0.5 N HCl aqueous solution and deaerated 1 N NaCl aqueous solution at

328 K and (b) deaerated 25 mass% NaCl aqueous solution at 333 K.

0

50

100

150

200

250

300

350

400

Cri

tica

l Pit

tin

g T

emp

erat

ure

, T

/ K

Ar-5 vol. % N2 Ar

SAF 2507 HDSS

Fig. 6 Effect of shielding gas composition on the critical pitting temper-

ature (CPT) of the HDSS welds and the commercial tube-SAF 2507 welds.


The PREN eq. (4) of DSS has been used by severalresearchers to investigate the resistance to localized corro-sion of duplex stainless steels.37,38)

When the PREN values are calculated by eq. (3) using anitrogen factor of 16, the PREN differences (PREN� �

PREN�) show negative values; thus, the PREN values of the�-phases are higher than those of the �-phases, except for theHAZ of the HDSS tube welded with an Ar shielding gas(Fig. 8(a)). Hence, it is reasonable to explain the differenceof corrosion resistance between these two phases through the

* Weld Metal: WM, Heat Affected Zone: HAZ

Fig. 7 Effect of shielding gas composition on the optical microstructures and the initiation and propagation of pitting corrosion of the

HDSS welds after a critical pitting test in 10 mass% aqueous solution of FeCl3�6H2O.

(a) BM

-2.0

-1.8

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4Welded with ArWelded with Ar-5 vol. % N2

PR

EN

(γ)−

PR

EN

( α)

WMHAZ(b)

BM0

1

2

3

4

5

6

7

8

PR

EN

(γ)−

PR

EN

(α)

Welded with ArWelded with Ar-5 vol. % N2

WMHAZ

Fig. 8 Effect of shielding gas composition on the PREN difference (PREN� � PREN�) between the �-phase and the �-phase of the HDSS

welds: (a) a nitrogen factor of 16 and (b) a nitrogen factor of 30.


PREN values calculated by eq. (4) using a nitrogen factor of30 (Fig. 8(b)) because pitting corrosion was initiated at the�-phases, and finally propagated from the �-phases to the �-phases (Fig. 7). Thus, based upon the PREN� and PREN�

values calculated by eq. (4) using a nitrogen factor of 30, thepitting corrosion must selectively generate at the �-phases ina highly concentrated Cl� environment, because the PRENvalue of the �-phase is much larger than that of the �-phase,irrespective of the chemical compositions of the shielding gasand the zones of WM, HAZ and BM (Fig. 8(b)).

As presented in Fig. 3 and Fig. 8(b), the PREN differencebetween the two phases is proportional to the ferrite content.As the content of the �-phase increases and that of the�-phase decreases, Cr, Mo, and W, which act as the �-stabilizers, are diluted in the �-phase and are enriched in the�-phase. In contrast, nitrogen, �-stabilizer, is restricted to amaximum of 0.05 mass% due to a small interstitial site of theBCC structure in the �-phase and is almost enriched in the �-phase. These results are well presented in Table 2.

Figure 9 presents the effects of shielding gas and factors ofalloying elements in PREN on the resistance to pittingcorrosion at the WM and HAZ of the HDSS welds. In thecase of the N2 supplemented Ar shielding gas, the ferritecontent decreases whereas the austenite content increases.

Table 2 Effects of shielding gas, volume fraction of � and �, and Nitrogen factor in PREN formula on the balance mechanism of the

resistance to pitting corrosion between austenite and ferrite in HDSS welds.

PRENN

ConditionChemical Compositions

%Cr + 3.3(%Mo + 0.5%W) + 16 or 30%N(mass%)

Nitrogen Factor of 16 Nitrogen Factor of 30

Cr Mo W N PREN �PREN PREN �PREN

substrate (� þ �) 27.00 2.57 3.23 0.34 46.3 PRENð�Þ � PRENð�Þ 51.0 PRENð�Þ � PRENð�Þ

�28.20 2.98 3.77 0.05 45.0 45.8

WZ(59 vol%) �PREN ¼ �0:2 �PREN ¼ þ7:1

�25.17 1.94 2.40 0.58 44.8

PRENð�Þ < PRENð�Þ52.9

PRENð�Þ > PRENð�Þ(41 vol%)

�28.12 2.96 3.75 0.05 44.9 45.6

Ar HAZ(63 vol%) �PREN ¼ þ0:4 �PREN ¼ þ8:3

�25.19 1.94 2.41 0.61 45.3

PRENð�Þ > PRENð�Þ53.9


�28.32 3.04 3.86 0.05 45.5 46.8

BM(51 vol%) �PREN ¼ �2:0 �PREN ¼ þ4:4

�25.13 1.91 2.37 0.51 43.5



�28.26 3.02 3.83 0.05 45.3 46

WZ(56 vol%) �PREN ¼ �1:2 �PREN ¼ þ5:7

�25.16 1.93 2.38 0.54 44.1



Ar �28.21 3.00 3.80 0.05 45.2 45.9

+HAZ

(60 vol%) �PREN ¼ �0:3 �PREN ¼ þ7:3

5% �25.17 1.93 2.40 0.59 44.9


PRENð�Þ > PRENð�ÞN2 (40 vol%)

�28.30 3.04 3.87 0.05 45.5 46.2

BM(50 vol%) �PREN ¼ �2:0 �PREN ¼ þ4:4

�25.14 1.90 2.37 0.51 43.5



(a)

(b)

Cr

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0.0

0.1

FAE

P1)

×[M

Ar

+ 5v

ol.%

N2

– M

Ar]

2)

Alloying Elements

Austenite Ferrite

Cr

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

FAE

P1)

×[M

Ar

+ 5v

ol.%

N2

– M

Ar]2

)

Alloying Elements

Austenite Ferrite

30N16NWMo

30N16NWMo

1) FAEP: factor of alloying elements in PREN, 2) [MAr+5Vol.%N2 – MAr]: the difference of chemical composition of each element between same phases.

Fig. 9 Effects of shielding gas and nitrogen factor in PREN on the

resistance to pitting corrosion at the WM and HAZ of the HDSS welds:

(a) HAZ (Heat Affected Zone) and (b) WM (Weld Metal).


Thus, Cr, Mo, and W are enriched in the �-phase whereasnitrogen is diluted in the �-phase, compared with Arshielding gas. The difference of chemical composition ofeach element between same phases is in the range of0:01�0:09 mass% and is very small (Table 2). However,the factor of Nitrogen in PREN is much larger than that of Cr,Mo and W. Accordingly, nitrogen has the most powerfulinfluence on the balance of corrosion resistance between the�-phase and the �-phase, compared with Cr, Mo, and W.

Figure 10(a) and Fig. 10(b) show the schematic of theeffect of the shielding gas composition on the phasetransformation in the weld metal of the HDSS welds. Aspresented in Fig. 10(b), the �-phase fraction in the weldmetal after welding with N2 supplemented Ar shielding gaswas decreased due to the addition of nitrogen that increasesthe transformation temperature from the �-phase to the �-phase during cooling period. As a result, Cr2N was alsodecreased due to nitrogen’s increasing solubility associatedwith the increasing amounts of austenite. Hence, theresistance to pitting corrosion was increased after welding

with N2 supplemented Ar shielding gas, compared with thatof Ar shielding gas. Figure 10(c) show the mechanism of thepitting corrosion in the weld metal of the HDSS welds.As mentioned in Fig. 4(d), a Cr-depleted zone decreases thecorrosion resistance due to a small Cr concentrationcompared with that of the matrix. Also, the pitting corrosionwas selectively initiated at the �-phase, and was thenpropagated from the �-phase to the �-phase (Fig. 7). Hence,the order of priority of pitting corrosion resistance are asfollows; �-phase > �-phase > Cr-depleted zone.

In summary, the reasons that the resistance to pittingcorrosion of the HDSS welds decreases are as follows; first,the difference of the PREN between the �-phase and the �-phase increases with an increase of the �-phase. Second, theCr2N that causes the Cr-depleted zone increases with anincrease of the �-phase.

Therefore, to decrease the PREN difference between thetwo phases and thereby increase the corrosion resistance,the ferrite content should be decreased. In particular, theaddition of N2 gas as a powerful �-stabilizer to the shieldinggas helps to decrease the ferrite content in the weld metal ofthe HDSS welds.

4. Conclusions

To elucidate the effect of shielding gas composition on thephase transformation and the mechanism of pitting corrosionof hyper duplex stainless steel (HDSS) welds, a metallo-graphic examination, a potentiodynamic polarization test, acritical pitting temperature test, scanning electron micro-scope and energy dispersive spectroscope (SEM–EDS)analyses, a scanning Auger multi-probe (SAM) analysis ofthe �-phase and the �-phase, and a transmission electronmicroscope (TEM) analysis of Cr2N precipitates were carriedout. From the results of these tests the following conclusionshave been drawn.

(1) During cooling after welding the HDSS tube usingGTAW method, a low heat input and a high cooling ratecaused an increase of �-phase and Cr2N in the weld metal andthe heat affected zone, whereas they suppressed precipitationof intermetallic compounds such as the �-phase and �-phase.

(2) Cr nitrides (Cr2N) were precipitated in the weld metaland heat affected zone due to the high �-phases content. Cr-depleted zone adjacent to Cr2N decreased the resistance topitting corrosion.

(3) It was verified that the resistance to pitting corrosionof the HDSS tube after welding with N2 supplemented Arshielding gas was increased due to a decrease of �-phase inthe weld metal and heat affected zone. The improvedcorrosion resistance is also attributed to a decrease of thePREN difference between the �-phase and �-phase in theweld metal.

(4) Based on the PREN� and PREN� values calculatedusing an N factor of 30, pitting corrosion in the weld metaland heat affected zone in the HDSS tube after welding wasselectively initiated at the �-phases because the PREN valueof the �-phase is much larger than that of the �-phase,irrespective of the chemical compositions of the shieldinggas. The pitting corrosion was finally propagated from the�-phase to the �-phase.

Fig. 10 Schematic diagrams of the effect of shielding gas composition on

the phase transformation and mechanism of pitting corrosion in the weld

metal of the HDSS tube welds: (a) the phase transformation after welding

with shielding gas of pure Ar, (b) the phase transformation after welding

with shielding gas of Ar-5 vol% N2 and (c) the mechanism of pitting

corrosion.


(5) It was found that the resistance to pitting corrosion ofthe HDSS tube with PREN 51 is superior to that of a superduplex grade SAF 2507 with PREN 46 in highly concentratedchloride (Cl�) environment.

REFERENCES

1) J. Olson and S. Nordin: Proc. Duplex Stainless Steel ’86, (1986)

pp. 219–225.

2) E. Perteneder, J. Tosch, P. Reiterer and G. Rabensteiner: Proc. Duplex

Stainless Steel ’86, (1986) pp. 48–56.

3) J.-O. Nilsson: Mater. Sci. Technol. 8 (1992) 685–700.

4) The International Molybdenum Association (IMOA): Practical Guide-

lines for the Fabrication of Duplex Stainless Steel, (Pergamon Press,

2009) pp. 1–64.

5) H. Okamoto: Proc. Applications of Stainless Steels ’92, Vol. 1, (1992)

pp. 360–368.

6) S. Atamert and J. E. King: Mater. Sci. Technol. 8 (1992) 896–911.

7) P. J. Ferreira and S. Hertzman: Proc. Duplex Stainless Steel ’91, Vol. 2,

(1991) pp. 959–966.

8) B. E. S. Lindblom, B. Lundqvist and N. E. Hannerz: Proc. Duplex

Stainless Steel ’91, Vol. 1, (1991) pp. 373–382.

9) H. L. Cao and S. Hertzman: Proc. Duplex Stainless Steel ’91, Vol. 1,

(1991) p. 363–372.

10) B. E. S. Lindblom and N. Hannerz: Proc. Duplex Stainless Steel ’91,

Vol. 2, (1991) pp. 951–958.

11) U. Draugelates, A. Schram, C. Boppert and J. Liu: Proc. Duplex


12) S. A. Urmston, G. K. Creffield, M. A. Cole and W. Huang: Proc.

Duplex Stainless Steel ’94, Vol. 2, (1994) Paper 27.

13) G. Rabensteiner and J. Tosch: Proc. Duplex Stainless Steel ’94, Vol. 2,

(1994) Paper 44.

14) R. Bradshaw and R. A. Cottis: Proc. Duplex Stainless Steel ’94, Vol. 2,

(1994) Paper 31.

15) B. Bonnefois, J. Charles, F. Dupoiron and P. Soulignac: Proc. Duplex


16) J. J. Dufrance: Proc. Duplex Stainless Steel ’91, Vol. 2, (1991) pp. 967–

975.

17) U. Draugelates, A. Schram and C. Boppert: Proc. Duplex Stainless

Steel ’91, Vol. 2, (1991) pp. 985–992.

18) R. Roberti, W. Nicodemi, G. M. La Vecchia and S. Basha: Proc.

Duplex Stainless Steel 91, Vol. 2, (1991) pp. 993–1000.

19) C. Boppert and U. Draugelates: Proc. Duplex Stainless Steel ’94, Vol. 3,

Paper 73.

20) J. C. Lippold, W. Lin, S. Brandi, I. Varol and W. A. Baeslack III: Proc.

Duplex Stainless Steel ’94, Vol. 1, (1994) Paper 116.

21) E. I. Kivineva and N. E. Hannerz: Proc. Duplex Stainless Steel ’94,

Vol. 1, (1994) Paper 7.

22) H. Hoffmeister and G. Lothongkum: Proc. Duplex Stainless Steel ’94,

Vol. 1, (1994) Paper 55.

23) J. C. Lippold, W. A. Baselack and I. Varol: Weld. J. 71 (1992) 1–14.

24) I. Varol, J. C. Lippold and W. A. Baeslack III: Key Eng. Mater. 69

(1992) 217.

25) G. G. Kolchin, B. S. Ermakov, R. I. Grechin and Y. V. Gladnev: Steel

USSR 17 (1987) 235–236.

26) L. A. Norstrom and S. Pettersson: Z. Werkstofftech. 12 (1981) 229–

234.

27) L. Weber and P. J. Uggowitzer: Proc. Int. Symp. on High Performance

Steels for Structural Applications, (1995) p. 291.

28) H. Kokawa, T. Kuwana and J. Okada: Welding Int. 7 (1993) 384.

29) H. Matsunaga, Y. S. Sato, H. Kokawa and T. Kuwana: Sci. Technol.

Weld. Join. 3 (1998) 225.

30) R. N. Gunn and P. Anderson: Proc. Duplex Stainless Steel ’94, Vol. 2,

(1994) Paper 30.

31) G. Tabensteiner and J. Tosch: Proc. Duplex Stainless Steel ’94, Vol. 2,

(1994) Paper 44.

32) Annual Book of ASTM Standards, ASTM G 5.

33) Annual Book of ASTM Standards, ASTM G 48-method A.

34) K. Lorenz and G. Medawar: Thyssen Forschung 1 (1969) 97–108.

35) S. Bernhardsson: Proc. Duplex Stainless Steel ’91, Vol. 1, (1991)

pp. 185–210.

36) S. Herzman, W. Roberts and M. Lindenmo: Proc. Duplex Stainless

Steel ’86, (1986) pp. 257–262.

37) M. Barteri, M. G. Mecozzi and I. Nembrini: Proc. Duplex Stainless

Steels 94, Vol. 3, (1994) paper 60.

38) U. Heubner and M. Rockel: Werkst. Korros. 37 (1986) 7–12.


Effect of Shielding Gas Composition on Phase ... · PDF fileIn this work, to elucidate the eﬀect of shielding gas composition on the phase transformation and the mechanism of pitting

Documents